RE: Facebook's definition of a Place

In my mind we need to determine:

1. the Basic core attribution standard of a POI (name, location/anchor, contacts, type, UID, ...)
2. how to represent those core attributes (named attributes, fixed value types, ...)
3. via what diverse means of representation/notation (XML, RDF, JSON, ...)
4. how to open the model for ease of extension (av pair model, tags, references/associations,...)


_______________________________
Karl Seiler
Director Location Technology & Services
NAVTEQ - Chicago
(T)  +312-894-7231
(M) +312-375-5932
www.navteq.com


-----Original Message-----
From: public-poiwg-request@w3.org [mailto:public-poiwg-request@w3.org] On Behalf Of Thomas Wrobel
Sent: Thursday, November 04, 2010 9:19 AM
To: Jens de Smit
Cc: Dan Brickley; Marengo Marco; public-poiwg@w3.org
Subject: Re: Facebook's definition of a Place

I think the only developers needing to convert would be ones taking
data from one protocol into another.
Most would just be putting the data in the format suitable for their
protocol/media.

ie. Everyone is using the same "Lat/Lon/Alt" definition, to the same
spec - but sometimes it makes more sense to express the data in the
form of, say, xml. And other times it might make more sense in the
form of JSON, or over XMPP, or dozens of other possibility's based on
usage.

Ensuring everyone is at least talking about the same values makes
conversion a (relatively) trivial process when it is needed. But if we
restrict it to one notation, that rules out a lot of use's and people
developing for those use's would end up using a completely separate
set of values altogether.

Imho, we could almost just look at key/value pairs needed to express
the data, as that seems the most generic form of expressing stuff
which could later be adapted for specific notations where needed.


On 4 November 2010 14:56, Jens de Smit <jens.desmit@surfnet.nl> wrote:
> On 04/11/2010 14:52, Dan Brickley wrote:
>>
>>
>>
>>
>> On 4 Nov 2010, at 14:00, Marengo Marco <marco.marengo@telecomitalia.it> wrote:
>>
>>> Hi,
>>> you may have heard that Facebook opened its APIs for reading (ok, it was available since a few months) and writing (since yesterday). I think Facebook's definition of "POI" might be interesting for our WG, even if it looks a bit too simple to me.
>>
>> Yes, imho this is very relevant. See also their http://opengraphprotocol.org which has some basic location pieces too, in RDFa. I think dealing with multiple notations will be inevitable for this wg.
>
> One thing I'm worried about when using multiple notations is that, as a
> user/developer you will need tools to convert between one notation and
> the other. Won't this be very obfuscated to the uninitiated and thereby
> prevent adoption?
>
> Regards,
>
> Jens
>
>



The information contained in this communication may be CONFIDENTIAL and is intended only for the use of the recipient(s) named above.  If you are not the intended recipient, you are hereby notified that any dissemination, distribution, or copying of this communication, or any of its contents, is strictly prohibited.  If you have received this communication in error, please notify the sender and delete/destroy the original message and any copy of it from your computer or paper files.

Received on Thursday, 4 November 2010 14:28:31 UTC